Summary: 18:28utc: Out of OBSERVING (CORRECTIVE_MAINTENANCE) for PSL refcav alignment
Sheila checked on the high freq Intensity noise reported by Georgia last night, and has traced this down to PSL Ref Cav transmission (talked with Jason & heard him mention he has noticed this transmission dropping when humidity in the H1 PSL room rises). Consulted with Keita & we decided to take H1 out of OBSERVING & mark this time as CORRECTIVE_MAINTEANCE (FRS forthcoming). Keita contacting Brian & I notified other observatory operators via TeamSpeak.
RefCav alignment touch up is being done by Jason here in the control room via picomotors on the table. This touch up caused a lockloss. Once Jason is complete, I will take H1 to DC Readout and then we will take this opportunity to check H1's ability to acquire (I assume this will involve alignment stuff---i.e. ASC & Dither line systems).
The highest I could get the RefCav TPD voltage was ~2.56V; this is in contrast to our usual max of ~4.2V. I'm not sure the reason for this difference, but it could be temperature/humidity related. I took a quick look at temperature and humidity in the laser enclosure over the last few days, see attachment. The swings in temp and %RH seen in the first half of the plot represent an incursion into the PSL enclosure this last Tuesday, 5/14/2019. The table north temperature sensor has seen an increase since we left the enclosure, from ~74 °F to ~74.5 °F, while the table south temperature sensor has seen no change. The humidity in the laser room has seen a large increase since the last incursion, it's currently sitting at ~42% (this makes sense as we've had some spring rain move into the area the latter half of this week).
Once we got to NLN the high frequency noise in DARM was still bad.
Keita adjusted the FSS common gain (H1:PSL-FSS_COMMON_GAIN) from 20 dB to
24.5 dB23 dB, which got rid of the noise. We tried to find a place in the guardian where this gain is set but it looks like it isn't hard coded anywhere, and it will just go back to its original value.If we lose lock and re-acquire and this gain is somehow reverted to 20dB, operators could try stepping it back (eg by running the command
z step H1:PSL-FSS_COMMON_GAIN +0.1until it reaches 23).The change was quite drastic, a difference of 1dB or so suddenly quenched terrible high frequency stuff.
24.5dB was based on the fact that the last time FSS common gain was set to 20dB was when the refcav transmission was 4.2, now it is 2.5, so it was the reduction of 4.5dB.
However, after doing that we noticed that H1:PSL-FSS_FAST_MON became quite glitchy, so I reduced it to 23dB, which looks OK.